Kimi-K2.6-Reasoning-3300x-WandB is a synthetic reasoning dataset containing 3,303 accepted rows. It was generated using the Kimi-K2.6 model through Weights & Biases Inference and is a subset of a larger planned 8,000-example distillation run. The dataset was created by author trjxter and last updated on 2026-05-18.
Use Cases
- Training or fine-tuning language models for reasoning tasks based on synthetic examples.
- Benchmarking the reasoning capabilities of language models based on generated data.
- Studying the properties and quality of model-generated synthetic data for distillation.
Strengths
- Contains 3,303 accepted rows of generated reasoning data.
- Data was audited and cleaned after generation, as mentioned in the description.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is limited to a subset of a larger planned run, as generation stopped due to a quota limit.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- trjxter on Hugging Face
- Collection Method
- Synthetically generated using the Kimi-K2.6 model via Weights & Biases Inference.
- Freshness
- Last updated 2026-05-18 03:28:01; freshness should be verified.